INDEX

Explanations

expressions and phrases indicating personal preferences or favorites

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 שוליים

-0.82

ſammen

-0.77

 queſta

-0.71

NameInMap

-0.69

ſelben

-0.66

 Anſ

-0.66

 للاسماء

-0.63

 raiſ

-0.62

Diwedd

-0.62

 ſen

-0.61

POSITIVE LOGITS

 favorite

0.73

 favourite

0.63

 favorites

0.57

favorite

0.56

 Favorite

0.53

 FAVORITE

0.53

Favorite

0.50

 favorit

0.49

 ulub

0.47

 preferred

0.47

Activations Density 0.010%

expressions and phrases indicating personal preferences or favorites

No Comments

No Known Activations